Loss of separation

FACTUAL INFORMATION

An Airbus Industrie A340 registered as BHXI and operating flight number Cathay 104 (CPA104) had been flight-planned to operate a flight from Melbourne to Hong Kong. The crew of the A340 had been cleared to depart Melbourne on a KEPPA TWO standard instrument departure (SID) with a requirement to maintain 5,000 ft. The KEPPA TWO departure was able to be conducted from Melbourne runways 16, 27 or 34. The Melbourne Automatic Terminal Information service (ATIS) was "Information Sierra" with a variable easterly wind at 5 kts. Downwind on runway 27 was reported as 5 kts. The duty runway at Melbourne was runway 16 for arrivals and runway 27 for departures.

The departure of CPA104 was coordinated with both the Departures North controller and the Essendon aerodrome controller (ADC) by the Melbourne ADC. The crew of CPA104 was cleared for take-off and instructed to contact Departures airborne.

A Cessna C172 registered as VH-ZWR had been flight-planned from Essendon to Kyneton at 3,500 ft, flying under the visual flight rules. The Essendon ADC had cleared the pilot in command to depart the Melbourne CTR on an amended route via Rockbank at an amended level of 1,500 ft. The aircraft had departed from Essendon's runway 17 and had made a right turn to track via Rockbank. Because the C172 would transit only controlled airspace that was the responsibility of Essendon Tower, coordination for this aircraft with other control agencies was not required.

The Departures North controller identified CPA104 airborne and cleared the crew to climb to flight level 200. The controller then observed on radar an unidentified aircraft squawking code 4000, departing Essendon westbound and crossing the departure track of the A340. The controller initiated corrective action but the A340 passed behind the unidentified aircraft with 1 NM horizontal and 900 ft vertical separation. The separation standard was infringed: the required standard between these two aircraft was either 3 NM or 1,000 ft. The unidentified aircraft was subsequently confirmed to be VH-ZWR.

Air traffic control coordination procedures required the Melbourne ADC to obtain departure instructions for CPA104 from the Departures North controller, as the aircraft was planned to depart on a northerly track. The coordination between ADC and Departures was carried out and an "unrestricted" clearance was obtained. The use of the non-duty runway for departures from Melbourne was not unusual. Runway 16 was often used for aircraft tracking to southerly destinations or for the larger international aircraft requiring the longer runway.

Because CPA104 was departing from runway 16, additional coordination was required between the Melbourne ADC and the Essendon ADC in accordance with Local Instruction LOA2976 - Coordination for Non-Duty Departures Runway 16. Coordination with the Essendon ADC was attempted; however, due to conflicting traffic, the departure was not authorised by the Essendon ADC at the time.

The workload at Essendon was considered to be high with controllers working in all operational positions. Additional airspace had been negotiated and released to Essendon by the Melbourne Centre. The airspace configuration on this day was unusual. The usual configuration was for Essendon Tower to have the south-east quadrant of the Melbourne control zone (CTR) up to and including 2,000 ft. The south-east quadrant was from south of the extended centreline to runway 26 to east of the extended centreline from runway 17. For a planned calibration of the Essendon instrument landing system (ILS), an extension to the airspace was agreed that would encompass normal airspace, plus the airspace from the extended centreline of runway 26 to the western edge of the CTR up to and including 3,000 ft. This additional airspace included all airspace south of the extended centreline of Essendon's runway 26 up to and including 3,000 ft.

The aircraft conducting the calibration testing of the ILS was operating to and overshooting from Essendon's runway 26. The aircraft was an Astra 1125 jet aircraft operating under the callsign of Auscal 01 (ADA01). This was the first time that this type of aircraft had been used for calibration tests: previously the Airservices Australia Fokker F28 had conducted the testing. When the F28 did the testing, it was able to overshoot from the approach prior to crossing the runway intersection, which obviated the need to sequence the aircraft with other traffic. The Astra aircraft needed to conduct an overshoot from all approaches, which necessitated complex sequencing with all other traffic. It was this aircraft that precluded CPA104 getting airborne when the Melbourne ADC first attempted to coordinate a departure with the Essendon ADC.

When the Melbourne ADC next attempted to coordinate with the Essendon ADC the departure of CPA104, the Essendon controller pre-empted the request as soon as the "hotline" communication line was opened by issuing a clearance using the word "approved" for the Melbourne runway 16 departure. Melbourne Tower advised that "he (CPA104) would be after the one on short final". The Essendon ADC replied with the statement "OK, behind him, approved". The Essendon ADC stated during interview that he couldn't recall issuing the approval for the departure of CPA104.

Although not a documented procedure, the practice in Essendon Tower was to place a runway-16 departure designator strip in the bay containing departure strips. The designator strip was used to remind and alert controllers of aircraft cleared to depart from runway 16 at Melbourne. Essendon Tower did not hold individual flight progress strips for Melbourne's aircraft and the runway-16 designator strip was used to indicate a pending departing aircraft. Controllers were trained to scan the strip bay to identify conflicting aircraft prior to issuing a subsequent clearance to an aircraft that could potentially conflict with aircraft departing Melbourne's runway 16. The Essendon ADC stated during interview that although the designator strip was placed in the departure bay, he did not scan the strip bay prior to clearing VH-ZWR for departure and the right turn.

ANALYSIS

Although the Essendon ADC did not recall approving the departure clearance for CPA104, the designator strip was reputedly placed in the departure bay. Analysis of the audio recording indicated that the Essendon ADC issued the take-off clearance for VH-ZWR approximately 3 seconds after approving CPA104's departure. This action indicated that the Essendon ADC did not recognise the potential for conflict between the two aircraft when the judgement was made to clear VH-ZWR for take-off and make the right turn. An alternative hypothesis is that a judgement was made that there was no conflicting traffic for the departure of VH-ZWR, prior to issue of the take-off clearance and receipt of the hotline call from Melbourne. The controller's mind-set regarding VH-ZWR did not change because the implications of CPA104's departure did not register with the controller. Importantly, the potential for conflict was not recognised when the Essendon ADC approved CPA104's departure. The latter hypothesis is supported by the controller's failure to scan the strips prior to the take-off transmission being made because his mind-set was that VH-ZWR was clear for take-off.

The traffic density and complexity on this particular day was reported to be unusually high. This additional workload may have increased the cognitive demands on the controller, unbeknown to other members of the team. The other controllers were also busy and the ADC was reputedly a very experienced controller. These two factors may have been the reason that the performance of the ADC was not monitored more closely.

Analysis of the radar data indicated that when the 3-NM radar separation standard was infringed, vertical separation indicated that CPA104 was 500 ft below VH-ZWR and climbing. When the aircraft were at the same level of 1,300 ft, radar separation had reduced to 2.2 NM. When CPA104 passed behind VH-ZWR, vertical separation had increased to 900 ft and radar separation had reduced to 1.25 NM. A vertical separation standard of 1,000 ft was re-established when the aircraft were 1.4 NM apart.

SAFETY ACTION

The Airservices Australia investigating officer made the following two recommendations:

"1. Amend LOA 2976 to:

a. Give better examples of the approval process for Melbourne runway 16 departures; and

b. Promulgate usage and the process to be followed by Essendon tower of the runway 16 designator strip.

2. Consider releasing all control zone airspace to Melbourne and Essendon towers".

Melbourne ATC management agreed to implement Recommendation 1 at the next revision of Local Instructions. Management will consider Recommendation 2 following consultation with terminal control unit staff.

Four high-capacity air transport passenger aircraft were en route to Perth under the control of the Melbourne Sector 1 controller, using procedural control methods. Two aircraft were on air route L513 and the other two aircraft were on air route Q76/J68. The routes converged and aircraft at the same level or not longitudinally separated would be in lateral conflict at the Perth 150 NM position for the routes. The aircraft were occupying FL270, FL280 and FL290. Aircraft on the same routes were separated by the minimum vertical separation standard of 1,000 ft.

However, the aircraft's estimates for the 150 NM Perth positions were within an eight minute period which did not provide a longitudinal separation standard. The controller elected to descend VH-TAX, a Boeing 737, which was at FL280 on route Q76/J68 to FL260 prior to the 150 NM position, to maintain separation with an aircraft at the same level on route L513. The crew descended to FL260 as instructed. Subsequently, the sector 1 controller was queried by another controller as to the separation standard used to maintain separation with VH-HYJ, an Airbus A320, which was on route Q76/J68 at FL270. The controller had descended TAX through the level of HYJ without an appropriate longitudinal standard being applied between the aircraft. There was a breakdown of separation.

The lack of a longitudinal standard between TAX and HYJ had been recognised by the previous sector controller. This controller had annotated the flight progress strips for HYJ to provide a prompt for distance checks. The sector 1 controller was nearing the end of his shift, during which he had worked the last three hours alone. To accommodate training commitments for other controllers and for his own purposes the controller's shifts had been swapped on a number of occasions. The number of recent shifts worked, the time of day and the fact that the controller was nearing the end of his shift probably combined to provide an environment in which he became less vigilant in the scanning of the flight progress strips. Consequently, he did not appreciate that there was not a longitudinal separation standard between the aircraft.

The Departures North controller was controlling a Beechcraft Super Kingair B200C, radio callsign Alpha-Mike-Mike, and a British Aerospace BAe146 radio callsign Echo-Whiskey-Mike. Both aircraft had departed Sydney to the south-east. The crew of the Kingair were flying a radar departure, on a heading of 120 degrees, maintaining 5,000 ft. The crew of the BAe146 were flying a runway 16L COOKS ONE standard instrument departure (SID).

The Departures North controller confused the callsigns of the two aircraft and, although intending to turn the B200C, mistakenly instructed the crew of the BAe146 to turn left onto a heading of 350 degrees. The BAe146 was then cleared to leave 5,000 ft on climb to flight level 240. The combination of both instructions placed the BAe146 in the Approach South/Director controller's airspace and in potential conflict with a Boeing B737 on left downwind for runway 25 that had been assigned descent to 5,000 ft.

Both the Approach South controller and the Departures North controller recognised that the aircraft were in conflict as soon as the BAe146 started to turn onto the assigned heading and issued remedial instructions to both crews.

The required separation standard was either 1,000 ft vertically or 3 NM horizontally. Analysis of the radar data indicates that the B737 did not descend below 5,500 ft and that the BAe146 did not climb above 4,600 ft. The vertical difference of 900 ft occurred while the aircraft were less than 3NM apart and there was consequently an infringement of the separation standards.

The crew of the BAe146 sighted the B737 and were able to maintain visual separation.

The investigation revealed that the Departures North controller had worked six extra shifts on overtime or emergency duty in the previous 7 weeks, including an overtime shift the previous day. Fatigue was considered to be a contributory factor.

LOCAL SAFETY ACTION

Air Traffic Control Management at Sydney made two recommendations as a result of this incident:

"Recommendation 1. That strategies be put in place to limit the amount of overtime or extra duties worked by terminal control unit (TCU) staff. Should the present measures be withdrawn, other measures will be necessary".

"Recommendation 2. That the mode of operation, 16 Departures/25 Arrivals, be reviewed so as to permit two outbound radials to the north for different performance aircraft".

BASI SAFETY ACTION

As a result of this investigation and a number of similar occurrences, the Bureau of Air Safety Investigation conducted an investigation of systemic issues at the Sydney terminal control unit and issued report B98/90 on 18 August 1998. Nine recommendations were made in the report. The following two recommendations are considered relevant to this investigation.

R980159

The Bureau of Air Safety Investigation recommends that Airservices Australia reassess the human factor hazard analysis for both Stage One and Stage Two of the long-term operating plan (LTOP) safety cases so that the mitigating strategies applied to identified hazards adequately allow for the fundamental limitations of human performance. In reassessing this hazard analysis, BASI recommends that Airservices Australia seek the assistance of human performance expertise.

R980160

The Bureau of Air Safety Investigation recommends that Airservices Australia consider restructuring the current roster operating in the Sydney TCU to ensure that contemporary fatigue management research is translated into meaningful duty hour regulations. In any restructure of the roster, BASI recommends that Airservices Australia expand its absentee management program to include individuals who expose themselves to the risks of fatigue by participating in excessive amounts of overtime and/or emergency duty.

The aerodrome controller (ADC) at Perth tower had cleared the Boeing 737 to line up on runway 21. An Airservices Australia car was accompanying the airfield mower cutting grass at the runway's edge, approximately midway along the runway. The ADC was aware that the car was just within the flight strip but he was not aware of the existence of the mower. Only after he cleared the 737 to take off and it had begun rolling, did he become aware of the mower and its close proximity to the runway. He instructed the pilot of the 737 to cancel the take-off.

The surface movement controller (SMC), who had not heard the take-off clearance, became aware of impending confliction when the aircraft began rolling. He instructed the car to vacate the flight strip. The aircraft stopped approximately 600 m before the original position of the mower having rolled approximately 750 m. The mower was a small blue vehicle and was approximately 1500 m from the tower and would have been difficult to see from the tower. The ADC had taken over the position approximately 10 minutes prior to the incident. He reported that he could not recall being advised about the mower during the handover so the ADC was unaware of the mower's existence.

However, the "runway occupied" strip had been placed in the ADC's departure runway designator console but because the ADC did not post the aircraft's flight strip on the console before clearing the aircraft for take-off, he did not note that the mower's operation near runway. The manual of air traffic services (MATS) details a requirement that an aircraft not be issued with a take-off clearance until its flight strip has been placed under the "runway occupied" strip at the ADC console and the runway has been vacated. When the ADC issued the take-off clearance, he had assumed that the SMC had ensured that the runway was vacated although he had not coordinated with the SMC to ensure that such was the case.

The tower coordinator normally supervised the operations within the tower but he was pre-occupied at the time recording information for the automatic terminal information service (ATIS). He had, therefore, not noticed the incident developing until the aircraft had begun its take-off run. He reported that he then called to the ADC to stop the aircraft because there was a vehicle on the runway.

LOCAL SAFETY ACTION

Airservices Australia has issued a local instruction highlighting the necessity of full handover procedures and a requirement to address escort vehicles with their callsign and escorted company on all radio communications and during tower coordination.

The airport and associated airspace were being operated in accordance with Mode 7 of the Sydney Long Term Operating Plan (LTOP), in that runway 25 was being used for departures while runways 34L and 34R were being used for arrivals.

A Boeing 737 (B737), registration VH-CZV (CZV), had departed runway 25 at Sydney. When the aircraft was 6 NM north-west of Sydney airport, the crew was cleared to climb to flight level (FL) 280. When it was 10 NM north of Sydney, the Departures North controller instructed the crew to turn right onto a heading of 060 degrees. Shortly after, the controller recognised that CZV was not going to reach 9,000 ft in sufficient time to maintain separation with another B737, VH-CZK (CZK), which was on a LETTI 3 Arrival, standard arrival route (STAR) for runway 34R and maintaining 8,000 ft. The required separation standard was either 1,000 ft vertically or 3 NM horizontally.

The aircraft were approximately 3 NM apart when traffic information was passed to the crew of CZV. Subsequently, the crew of CZV reported sighting CZK and advised the controller that visual separation could be maintained. Analysis of the radar data indicated that a breakdown of separation had occurred when the lateral separation standard was infringed while the vertical displacement of the aircraft was 500 ft. The aircraft subsequently closed to within 1 NM, at which point 1,400 ft of vertical separation existed. The investigation revealed that the controller cancelled restrictions for CZV that had been imposed for the departure. The speed restriction was cancelled first, followed by the altitude restriction of 5,000 ft.

The altitude restriction would have assured separation with the track of the inbound conflicting aircraft (CZK). When the altitude restriction was removed, the controller relied on monitoring the flight paths of the aircraft and his ability to implement any necessary action to maintain separation. The controller was undergoing a familiarisation period under the supervision of a suitably rated controller. The controllers were distracted, from the monitoring role, by coordination activities with flight service and the control tower. As a result of this investigation and a number of similar occurrences, the Bureau of Air Safety Investigation issued report B98/90 which covered the systemic investigation into factors underlying air safety occurrences in Sydney Terminal Area airspace.

FACTUAL INFORMATION

The crew of an instrument flight rules (IFR) Beech 200 (B200) was conducting a base check which would entail a practice non-directional beacon (NDB)/distance measuring equipment (DME) approach to runway 12 at Williamtown aerodrome. The pilot under check (handling pilot) and the checking pilot briefed for an approach to the minima, followed by a missed approach to 1,500 ft and then join the circuit for circuit training. The missed approach was to be conducted with the aircraft configured to simulate asymmetric operation. While inbound to Williamtown, the handling pilot briefed the approach controller of their intentions and was subsequently cleared to make the NDB/DME approach. The handling pilot then advised the approach controller that, after the NDB, the aircraft would conduct a go around. The check pilot did not hear this transmission due to the aircraft's communications configuration, which prevented his hearing any air/ground transmission by the handling pilot.

Prior to the missed approach point the handling pilot commenced an asymmetric missed approach and reported to the aerodrome controller. The aerodrome controller (ADC) instructed the crew to go around, to maintain runway heading and to maintain an altitude not above 500 ft. The transmission from the ADC was broken and both pilots believed that the assigned altitude was 1,500 ft. The handling pilot readback the requirement to maintain runway heading but did not readback the altitude, which was contrary to Aeronautical Information Publication (AIP) procedures. The ADC did not challenge the lack of the readback of the altitude by the pilot, which was contrary to air traffic control procedures.

The ADC's intention was to limit the altitude of the B200 to establish 500 ft vertical separation with a formation of Macchis that was entering the circuit via the initial point. The Macchi pilots had sighted the B200 as they tracked from the initial point and pitched into the circuit. During the missed approach, at approximately 1,200 ft, the B200 pilots saw the formation of Macchis pass from their left to right in front of and slightly below the level of their aircraft. The Macchi formation passed with approximately 200 ft vertical separation.

ANALYSIS

The misunderstanding by the handling pilot of the B200 in relation to a "go around" and a "missed approach", and the fact that the check pilot did not hear the transmission, created a developing situation that was appreciated differently by the two pilots and the ADC. The minima for the "missed approach" was 570 ft, whereas a "go-around" could have been commenced from a lower altitude.

While the radio transmission quality made communication difficult, the lack of radiotelephony discipline by both the handling pilot and the ADC ensured that the opportunity to resolve any misunderstanding was lost.

SIGNIFICANT FACTORS

1. The pilot of the B200 requested a "go around" instead of a "missed approach"

2. The check pilot could not hear the handling pilot's air/ground transmissions.

3. The pilot did not readback the requirement to maintain 500 ft to the ADC.

4. The ADC did not challenge the pilot to readback the altitude requirement.

FACTUAL INFORMATION

Inbound to Launceston from Cape Barren, the pilot of a PA-23, VH-JSB, requested a 32L VOR/DME approach. Because of conflicting traffic, air traffic services (ATS) instructed JSB to enter the holding pattern, to maintain 7,000 ft, and report turning inbound in the holding pattern. The pilot acknowledged the instruction, but proceeded immediately into the 32L VOR/DME approach, descending from 7,000 ft. He reported turning inbound, but ATS believed the 'Inbound' call referred to JSB being inbound in the holding pattern. The pilot continued to fly the published 32L VOR/DME pattern and positioned the aircraft to intercept the 319 degrees radial as prescribed on the DAP chart. Minimum altitude specified before being established on the 319 degree radial is 2,800 ft.

Meanwhile, VH-OZV, a PA-31, was in cloud at 3,000 ft, conducting an ILS approach to runway 32L at Launceston. A company pilot occupied the right seat on the flight from Flinders Island. When passing the NILE locator, 7.7 NM DME distance, at 3,000 ft, the aircraft entered a cloud break and the right seat pilot sighted another aircraft about 200m ahead. The flying pilot took immediate action to avoid the other aircraft which was subsequently identified as JSB. JSB continued the 32L VOR/DME approach, and ATS provided OZV with separation.

ANALYSIS

The final approach paths for the runway 32L ILS and runway 32L VOR/DME are slightly different, with the ILS approach requiring the 313 degrees inbound radial to be flown, and the VOR/DME specifying the 319 degrees radial. The inbound turn in the VOR/DME pattern is a right turn to intercept the 319 degrees radial, and this requires the aircraft to fly across the inbound ILS approach track. An aircraft flying a VOR/DME approach would normally be inbound by 8NM, which is almost adjacent to the NILE locator, which aircraft utilise on an ILS approach.

Altitude requirements in the two approaches are also similar. At a distance of 7.7 NM DME, the distance of the NILE locator, an aircraft on an ILS approach would be descending from 3,000 ft, and an aircraft on a VOR/DME should be at about 2,900 ft.

Although aircraft may be carrying out different approaches to runway 32L, they need to be treated as if they are flying the same approach, and ATS apply separation standards accordingly. If a pilot misunderstands an ATS instruction, potential exists for two aircraft to be in close proximity.

The pilot of JSB misunderstood the ATS instruction to report inbound in the holding pattern, to mean to report inbound in the VOR/DME approach. Because of this misunderstanding, he commenced the approach and descended below 7,000 ft without a clearance. This action placed JSB in conflict with OZV which had been cleared to carry out an ILS approach to runway 32L. The 'Inbound' call to ATS did not provide a warning of the developing situation, as ATS were expecting the call in the holding pattern.

ATS had advised the pilot of JSB that the instruction to enter the holding pattern was because of other traffic in the area, including some making instrument approaches. Although adequate information was transmitted by ATS, the pilot's decision to descend from 7,000 ft and carry out the VOR/DME approach he originally requested, indicated a loss of situational awareness.

Further investigation showed that four minutes before the breakdown of separation between JSB and OZV, VH-IAM, a PA31, had departed from Launceston for Flinders Island. IAM was held at 6,000 ft to maintain separation from JSB which was thought to have been at 7,000 ft. However, when IAM departed, JSB had descended from 7,000 ft. With JSB descending to the southeast of the VOR and IAM climbing to the northeast, there was probably no breakdown of separation, but there was no separation assurance.

SIGNIFICANT FACTORS

1. JSB descended from 7,000 ft and commenced a 32L VOR/DME approach without a clearance.

2. In a cloud break, the right seat pilot of OZV sighted JSB.

3. The flying pilot in OZV took action to avoid JSB.

The Boeing 737 (B737) was maintaining flight level (FL) 350 and the Melbourne Central controller had completed all the required tasks involving that aircraft. Although the aircraft had not yet reached the point at which it would normally be handed off to the next sector (Sector 4), the controller decided to proceed with an early hand-off which was accepted by the Sector 4 controller. The Airbus A320 was maintaining FL370 on a track which crossed that of the B737. The appropriate vertical separation standard of 2,000 ft was being maintained. As the tracks of the aircraft converged, the crew of the A320 was given descent to FL350 by the Melbourne Central controller, who momentarily forgot about the B737.

As soon as the crew of the A320 reported leaving FL370, the controller realised that he could not guarantee the required horizontal separation standard of 5 NM and immediately instructed them to return to FL370. The A320 had descended to FL368 before the crew were able to arrest the descent and commence climb back to FL370 and re-establish a vertical separation standard. The investigation revealed that the aircraft were at the same flight level, with a horizontal separation of 4.7 NM, resulting in a breakdown of separation standards. The practice of early hand-off is being reviewed by Airservices Australia as it can lead to a reduced level of awareness by controllers.

The incident occurred on the morning Class E Radar airspace was introduced. Class E Radar airspace included defined airspace within radar coverage over NSW, from south of Ballina to north of Canberra from 8,500 ft to Flight Level (FL) 125. It excluded existing Class C, Class D and military restricted airspace. To operate in Class E Radar airspace, Instrument Flight Rules (IFR) flights required an airways clearance. Air traffic control provided separation from other IFR traffic.

VH-GKB, on an IFR flight, was cruising at 9,000 ft, and reported overhead Williamtown at 0731 ESuT, estimating Taree at 0754. VH-SMH, also on an IFR flight, departed Williamtown at 0732 tracking direct to Point Lookout, and climbing to planned level FL 250. At 0733.40, the flight service officer coordinated the departure of SMH to Brisbane Control. At 0734.28, the flight service officer instructed SMH to transfer to the Sector 15C frequency, and suggested that SMH maintain 8,000 ft due to IFR traffic at 9,000 ft. Sixteen seconds later, SMH contacted Brisbane Control, advising an altitude of 8,800 ft. The controller immediately advised the pilot that clearance was not available, and instructed him to descend to 8,000 ft. At this time, the indicated altitude of the aircraft on the controller's radar display was 8,700 ft. The horizontal distance between the aircraft was 4.7 NM. The minimum separation standard required was 5 NM horizontally, or 1,000 ft vertically.

The operator advised that the aircraft was operating at low weight. This resulted in a high rate of climb. The crew was aware of the lower limit of the Class E Radar airspace and expected to receive a clearance before the aircraft reached 8,500 ft. In the event, the expectation of a clearance combined with the high rate of climb of the aircraft resulted in insufficient anticipation being applied to level the aircraft at 8,500 ft. Replay of the recorded radio transmissions revealed that the transmission from the flight service officer suggesting that the aircraft maintain 8,000 ft was broken, incomplete and slightly garbled. Consequently, the crew did not hear the suggestion to maintain 8,000 ft.

Prior to the occurrence, company practice when climbing into controlled airspace had been to set the lower limit of the airspace (in this case 8,500 ft) pending receipt of an airways clearance. However, this provided only 500 ft separation from traffic cruising at 9,000 ft in Class E Radar airspace and advice from ATC was that traffic information is not passed in such situations. Based on this information, and the circumstances of the incident, company procedures were amended to set 8,000 ft in the assigned altitude indicator.

VH-HYG, an Airbus A320, was en route from Sydney to Perth at flight level (FL) 350. VH-TAG, a Boeing 737, was en route from Adelaide to Perth on the same route and was maintaining FL330. The aircraft position reports for FILET, a reporting point over the Great Australian Bight, revealed that there was a 6 minute interval between the aircraft with HYG in front. The crew of TAG requested approval to climb to FL350. The controller believed that the groundspeed of the aircraft over a 6 minute period would provide the necessary 30 NM longitudinal separation standard and instructed the crew to climb to FL350.

The controller requested both crews to report their distance from FLAKE, the reporting point west of FILET. The reports from the crews revealed that the distance between the aircraft was 16 NM. The required separation was either 30 NM or 2,000 ft vertically. There was a breakdown of separation standards. Subsequently, the controller instructed the crew of TAG to descend to FL330 to maintain separation. Controllers were required to check for distance standard between aircraft that had less than 10 minutes between their respective estimates. There was a considerable number of aircraft being managed by the controller at the time of the occurrence.

Following the occurrence the controller was replaced at the position. However, prior to the occurrence the controller had approved the crew of VH-HYE, an Airbus A320 en route from Perth to Adelaide, to climb from FL370 to FL390 but had not coordinated this change of level with the next control position. Additionally, the controller had not annotated the new level on the HYE flight progress strip for HYDRA. This flight progress strip would be used to coordinate the position of HYE with the next control position. A second handover/takeover was then conducted with a third controller assuming responsibility for the position. The crew of HYE reported at both the HITCH and HOLLA positions at FL390 while this controller was operating the position. The controller advised Adelaide Sector 4 of HYE's HYDRA estimate and level from the flight progress strips for that position. The crew of HYE reported to Adelaide Sector 4 at HYDRA at FL390.

The level was confirmed by Sector 4 from the radar display label. The investigation established that the first controller may have been distracted by aircraft radio transmissions and coordination calls from other control positions while he was endeavouring to annotate all flight progress strips for HYE with the new level. Consequently, he did not annotate the new level, of FL390, in the HYDRA flight progress strip. In relation to TAG, the investigation did not establish the reason for the controller not complying with the requirement to check the distance between aircraft. In both handover/takeovers it is probable that the on-coming controllers did not conduct an adequate review of the flight progress strips. To some degree they may have been limited in their ability to conduct a check of the FPSs because of the level and complexity of the traffic situation. This aspect may also be the reason for the third controller not establishing the difference between the levels provided in the two position reports and that annotated on the HYDRA flight progress strip which was subsequently used for the coordination with Adelaide.